Objective To evaluate the cl inical results of allogeneic bone graft for interbody fusion in cervical tuberculosis. Methods Between January 2000 and January 2008, 30 cases of cervical tuberculosis were treated with allogeneic (group A, n=15) or autologous (group B, n=15) il iac crest bone graft combined with anterior fixation after radical debridement. In group A, there were 8 males and 7 females with an average age of 38 years; the disease duration was 6 to 14 months; the preoperative kyphosis Cobb angle was (8.6 ± 11.3)°; the preoperative Japanese Orthopaedic Association (JOA) score was 13.0 ± 3.1 for neurological function; and the length of bone graft was 32 mm on average. In group B, there were 9males and 6 females with an average age of 42 years; the disease duration was 4 to 17 months; the preoperative kyphosis Cobb angle was (4.9 ± 7.4)°; the preoperative JOA score 12.3 ± 4.2; and the length of bone graft was 34 mm on average. There was no significant difference in general data between 2 groups (P gt; 0.05). Results The operation time and bleeding volume in group A were significantly less than those in group B (P lt; 0.05). Wound effusion were found in 2 cases of group A, and the other incisions healed by first intention. No infection occurred in group B. In group A, 13 cases were followed up 12-48 months; in group B, 14 cases were followed up 13-46 months. The time of bone graft heal ing in group A [(7.6 ± 2.1) months] was significantly longer than that in group B [(4.2 ± 1.1) months] (t=2.773, P=0.005). The kyphosis Cobb angles were significantly improved at 6 months and last follow-up after operation in 2 groups when compared with that before operation (P lt; 0.05), but no significant difference was found between 2 groups at different time after operation (P gt; 0.05). There was no significant difference in JOA score at 6 months after operation between group A (14.1 ± 2.6) and group B (14.3 ± 2.4) (t=1.655, P=0.162). The improvement rate for neural function were 83.7% in group A and 87.8% in group B, showing no significant difference (χ2=3.150, P=0.071). There was no loosening of internal fixation and recurrence of tuberculosis in 2 groups during follow-up. Five cases had chronic pain at il iac donor sites in group B. According to Bridwell et al. evaluation standard, the bone fusion was satisfactory in 11 cases (84.6%) and unsatisfactory in 2 cases (15.4%) in group A, and was satisfactory for all in 14 cases (100%) in group B. The satisfactory rate of bone fusion showed no significant difference between 2 groups (χ2=2.680, P=0.115).Conclusion Allogeneic bone grafting has a good cl inical result for spinal fusion in cervical tuberculosis surgery, which can treat tuberculosis bone defect effectively.